Yeah this article makes no sense, really seems like puff piece slapped together in 15 min from a PR email he sent them. Retired 36 year old multi millionaire shows people they too can solve math puzzles for the low price of 2 mill?
using only publicly available unused cloud storage space … “I was able to do it just by using big tech’s leftovers.”
So he ostensibly used donated storage? Is computing primes even storage intensive? What about the compute? Is that where the 2 million went?
Plus what would “publicly available unused space” even be? Cloud providers don’t just give stuff away to random people because they aren’t using it
He used servers in 17 countries across 24 data centers and on two cloud providers to find the number,
So he was just running this on 2 big cloud services. Did the author just misunderstand him getting some credits from the providers? Did they run his workload as a low priority thing when they were over provisioned, is that what they mean by “leftovers”? Who knows.
he and his programmer father were writing a prime number computer code.
3,00o[sic] to 5,000 volunteers
Come on WashPo doesn’t have editors anymore?
Durant said he also spent the time and money on finding a prime number to show that GPUs can be used for more than AI
What a banal statement. Seems like it’s just a way to shoehorn in a way to bring up AI in the article.
Are you not aware that 99 times out of 100 if you see a tech company rapidly growing it’s completely unprofitable and not even attempting to be profitable yet? It’s called blitzscaling and is pretty clearly what openai is attempting. Like if you see a tech company quickly growing you should be assuming it’s unprofitable until proven otherwise not the opposite lol.